Live systems · Dubai corporate-services provider
I worked across a connected set of live customer and revenue workflows: web conversion, structured qualification, approved-knowledge retrieval, CRM and booking handoffs, referral operations, measurement and human escalation.
The working context
The problem
The individual pieces were not the difficult part. The difficulty was keeping the customer journey, commercial logic and internal follow-up consistent when several systems held a different part of the story. If one handoff lost context, the next team had to reconstruct it.
My contribution
I led product and workflow decisions across the consultation journey, conversational AI, qualification, CRM, scheduling, referral operations and measurement. I also worked through the less visible decisions: what the AI could say or do, when deterministic software should take over, and when a person needed the full context.
I have kept the organisation, product and people private. I also left out exact figures and internal details that could identify the work. The status above tells you how far the implementation actually went.
How I worked through it
If you are working on a similar system, these are the decisions I would examine before choosing tools or adding more automation.
I mapped what the customer and the internal team needed to decide at each point. That kept the workflow focused on useful progress instead of collecting information without a purpose.
AI interpreted language and retrieved approved knowledge. Conventional software handled known rules, records and handoffs. People stayed responsible for uncertain, sensitive or commercially important situations.
Qualification only helped if the booking, CRM record and follow-up reflected what the customer had already shared. I treated that continuity as part of the experience, not back-office plumbing.
Live workflows fail in ordinary ways. Messages arrive twice, state changes and integrations time out. I planned for retries, visible ownership and a route back to a person.
Expert insights from the work
These are the lessons I took from the work. I would use them as questions for your own implementation, not as a universal recipe.
Your conversational layer can answer well and still leave the work unfinished. Connect it to a controlled booking, CRM or human workflow that carries forward what the customer meant.
Put eligibility, ownership, consent and progress in structured state. Let the conversational layer interpret and explain, but do not ask it to be the only record of what happened.
When you escalate, preserve the reason, the relevant context and the next expected action. A transcript alone leaves the receiving team to work out why the handoff happened.
What I would improve next
My next pass would simplify where state lives and make operational visibility more explicit. The system should tell your team what happened, why a route was chosen and what needs attention next. I would spend effort there before trying to make the conversation sound cleverer.
Connected disciplines
Continue through the work
If this overlaps with a system you are thinking through, start a relevant conversation.
